Binomial Distribution Flashcards
Binomial distribution
a probability distribution that results when:
- there is a series of N trials
- on each trial, there are only two possible, mutually exclusive, outcomes
- there is independence between the outcomes of each trial
- the probability of each possible outcome on any trial stays the same from trial to trial
Binomial table
shows all possible outcomes and associated probabilities for any binomial distribution up to N=20
if P is higher number than represented on table, reverse it (ie, if P=.75, look at .25)
What are the two hypotheses that can be formulated for any observation?
the null hypothesis
the alternate hypothesis
mutually exclusive and exhaustive
Null hypothesis
H0 (0 is subscript)
the observation is due to chance or random factors
- more conservative, baseline assumption
Alternate hypothesis
H1 (1 is subscript)
the observation is not due to chance or random factors
- evidence to support H1 often suggests something new has been discovered or demonstrated
- if you do not accept H1, you “retain” or “fail to reject” H0
(we cannot conclude… or, this experiment does not provide evidence that…)
Alpha level (α)
an evidentiary standard set by the researcher before the study; defines how strongly the sample evidence must contradict the null hypothesis before you can reject it for the entire population
affects the probability of Type I and Type II errors
- more strict –> reduce chance of Type I (false alarm), but increase chance of Type II (miss)
decision rule
clear cutoff given by α (alpha level) for whether to reject or retain H0 (null hypothesis)
- if the obtained probability ≤ α, REJECT H0
- if the obtained probability > α, RETAIN H0
Normal approximation approach
with large numbers, normal distribution can be used to approximate binomial probabilities
Type I error
a decision to reject H0 when in fact H0 is true
- false alarm
Type II error
a decision to retain H0 when in fact H0 is false
- a “miss”
Type II error
a decision to retain H0 when in fact H0 is false
- a “miss”
Directional hypothesis
one-tailed hypothesis that states the direction of the difference or relationship
- only use if you have strong justification for formulating directional alternate hypothesis
Non-directional hypothesis
two-tailed hypothesis predicts that the IV will have an effect on the DV, but the direction of the effect is not specified
- more conservative approach